Abstract
Pancreatic fat accumulation is associated with age, obesity, and metabolic abnormalities. The contributions of individual metabolic and inflammatory measures after mutual adjustment remain unclear. Objective: To assess associations of a calculated semiquantitative pancreatic MRI fat index with clinical, anthropometric, metabolic, and inflammatory measures in adults with met-abolic risk factors. Materials and Methods: This cross-sectional analysis used prospectively collected data from 150 adults aged 30–60 years examined in Astana between 25 October 2025 and 30 April 2026. The MRI fat index was calculated from dual-echo chemical-shift images and was not equivalent to standardized MRI-PDFF. Analyses included Kruskal–Wallis tests, Spearman correlations, multivariable gamma regression, and sensitivity analyses. Multiple comparisons were adjusted using the Benjamini–Hochberg method. Results: The median index was 5.70 [2.60–9.83] %. The strongest positive correlations were with age (ρ=0.445), BMI (ρ=0.426), HbA1c (ρ=0.400), HOMA-IR (ρ=0.385), interleu-kin-6 (ρ=0.368), and insulin (ρ=0.354); HDL cholesterol was inversely correlated (ρ=−0.240). In the full gamma model, only age remained statistically significant, with a mean ratio of 1.510 per standard deviation (95 % CI 1.254–1.818; p<0.001). Associations with sex, BMI, and interleukin-6 reached statistical significance only in the additional linear model. Conclusion: In unadjusted analyses, the MRI fat index was associated with several an-thropometric, metabolic, and inflammatory measures. After mutual adjustment, the association with age was the most consistent. The semiquantitative nature of the index and the depend-ence of other associations on model choice limit the clinical interpretation of these findings